Testing for Racial Bias in Police Traffic Searches

I develop a framework to detect and measure racial bias in police traffic searches amid sample selection and statistical discrimination. By modeling the search decisions stochastically, I allow the direction and intensity of bias to depend on the officer’s belief of how likely a driver carries contraband. Sharp bounds on various measures of intensity are derived. This framework also enables me to evaluate each officer separately, thereby allowing for unrestricted heterogeneity in officer search preferences and beliefs.